Fusarium head blight is a major disease of wheat; it alters the yield and grain quality. In a context of reduced use of pesticides and regulations on mycotoxins at harvest, ARVALIS in collaboration with the University of Piacenza is developing a model that aims to optimize treatment against F. graminearum, the main mycotoxins-producing species on wheat. This mechanistic model describes the sexual stage of the pathogen. A first prototype was developed based on climatic parameters. To understand better the model outputs and the influence of each of the model parameters on these outputs, a sensitivity analysis was performed through three methods: Morris, ANOVA and Random Forest. These methods gave consistent results, enabling us to identify the most influential parameters.
Corre, C., Gourdain, E., Guénolé, G., Héritier, E., Manstretta, V., Rossi, V., Un modele épidémiologique pour piloter le traitement contre F. graminearum sur céréales à paille: analyse de sensibilité sur les paramètres climatiques, Paper, in AFPP – ONZIÈME CONFÉRENCE INTERNATIONALE SUR LES MALADIES DES PLANTES. TOURS – 7 AU 9 DÉCEMBRE 2015, (Tours, 07-09 December 2015), N/A, Tours 2015: 1-10 [http://hdl.handle.net/10807/69959]
Un modele épidémiologique pour piloter le traitement contre F. graminearum sur céréales à paille: analyse de sensibilité sur les paramètres climatiques
Manstretta, Valentina;Rossi, Vittorio
2015
Abstract
Fusarium head blight is a major disease of wheat; it alters the yield and grain quality. In a context of reduced use of pesticides and regulations on mycotoxins at harvest, ARVALIS in collaboration with the University of Piacenza is developing a model that aims to optimize treatment against F. graminearum, the main mycotoxins-producing species on wheat. This mechanistic model describes the sexual stage of the pathogen. A first prototype was developed based on climatic parameters. To understand better the model outputs and the influence of each of the model parameters on these outputs, a sensitivity analysis was performed through three methods: Morris, ANOVA and Random Forest. These methods gave consistent results, enabling us to identify the most influential parameters.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.